search

ABAtanasov / GalerkinSparseGrids.jl

Sparse Grid Discretization with the Discontinuous Galerkin Method for solving PDEs
Other
47 stars 14 forks source link

Upgrade to Julia 1.0 #4

Closed miguelraz closed 5 years ago

miguelraz commented 5 years ago

...There are many numerical relativity codes. There are many like it, but this one ain't mine...

Changes include:

I have all tests passing locally.

This PR can be squashed/rebased on top of #2 , but need some guidance on how to do so.

Also, there is a lot of enticing future work and optimizations to be worked for this package. Tapping into the full ecosystem of DifferentialEquations.jl, swapping into Tensors.jl, benchmarking and cleaning up the high speed criticality functions (like array2poly et al), applying @generated functions left and right, as well as revamping the code into a more Julian fashion (there's some keyword filtering everywhere, feels too... pythonic? Should exploit multiple dispatch.)

Let me know if more work is welcome, as I would love to dig into this. Great work!

ABAtanasov commented 5 years ago

Dear Miguel,

I'm glad this caught your eye, and thank you for these changes! I will take some time to look over them, but it looks like its up to speed with v1.0 . I'm glad that we're still using ODE.jl for now. I have a few technical things I need to first understand about the DifferentialEquations.jl before incorporating that ecosystem.

I am planning on adding one last major component to this package, involving the handling of nonlinear operations in sparse Galerkin bases. Adding this might very well restructure some large parts of the package. I expect this to all be done by December. After that, more work is certainly welcome and appreciated! @eschnett and I will keep you updated. In the meantime, I'll look at merging this by the end of the week.

miguelraz commented 5 years ago

Awesome! Most of the work was just grunt work of replacing 0.6 type signatures to 1.0 and swapping put deprecations. I can rebase on top of the DifferentialEquations.jlpackage - but yes, you will want the full experience ;) I'll keep posted then.

On 5 Nov 2018 20:13, "Alex Atanasov" notifications@github.com wrote:

Dear Miguel,

I'm glad this caught your eye, and thank you for these changes! I will take some time to look over them, but it looks like its up to speed with v1.0 . I'm glad that we're still using ODE.jl for now. I have a few technical things I need to first understand about the DifferentialEquations.jl before incorporating that ecosystem.

I am planning on adding one last major component to this package, involving the handling of nonlinear operations in sparse Galerkin bases. Adding this might very well restructure some large parts of the package. I expect this to all be done by December. After that, more work is certainly welcome and appreciated! @eschnett https://github.com/eschnett and I will keep you updated. In the meantime, I'll look at merging this by the end of the week.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/ABAtanasov/GalerkinSparseGrids.jl/pull/4#issuecomment-436107214, or mute the thread https://github.com/notifications/unsubscribe-auth/AMc4tRpY52cDUGbLfnTeXVcr2xEACByOks5usPBdgaJpZM4YBTpL .